GPT-5.5 API model
Reference for gpt-5.5 on InferGate: exact request ID, response validation, streaming behavior and account access troubleshooting.
Use gpt-5.5 as the exact request identifier. This is the currently published GPT-5.5 entry. Non-streaming and streaming Responses requests passed the production catalog acceptance on September 11, 2026.
Request this model
curl --fail-with-body https://api.useinfergate.com/v1/responses \
-H "Authorization: Bearer $INFERGATE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-5.5","input":"Reply with one sentence about API design.","max_output_tokens":256}'Choose deliberately
Compare model families using repeatable workloads. No benchmark, context-window or speed advantage is asserted here. Confirm required parameters using a small representative request.
Check access before sending requests
Use your own InferGate key to request the live catalog. Account permissions, key restrictions and available balance still apply. A model listed here is not a guarantee of uninterrupted upstream service.
curl --fail-with-body https://api.useinfergate.com/v1/models \
-H "Authorization: Bearer $INFERGATE_API_KEY"Retired model requests return model_not_available. InferGate does not silently substitute another model. Keep keys on your server and out of browser bundles.
Prepaid usage and costs
API usage consumes available account credit. Review current input and output rates in application pricing before running a workload. This page does not promise free requests or a fixed discount. See how prepaid pricing works and compare actual usage logs after a small test.
GPT-5.5 request reference
| Setting | Value | What to verify |
|---|---|---|
| Model | gpt-5.5 | Keep the full identifier, including any suffix |
| Base URL | https://api.useinfergate.com/v1 | Use the API host, not the website |
| Endpoint | POST /responses | Responses-format input and output |
| Modes | Complete response or stream=true | Require completed output in either mode |
Frequently asked questions
Which identifier should I send for GPT-5.5?
Send gpt-5.5 exactly. Use the same API key to check GET /v1/models before requesting this model. Display labels and family names are not replacements for the request ID.
Why can a published model be unavailable to my key?
The website describes the public catalog. Your account group, API key model restrictions and balance can affect a specific request. Inspect the actual error and authenticated list before retrying.
Does GPT-5.5 support every optional SDK parameter?
The verified baseline is text Responses requests in complete and streaming modes. Test any tool, file, modality or optional parameter you require separately; SDK support alone does not establish upstream support.
How do I verify a streamed response?
Require the completion event and useful text. Preserve the request ID and actual model identity. A connection that ends before completion is not a successful result merely because HTTP status was 200.
Continue your integration
- Start a GPT-5.5 API integration
- Handle streaming completion and interruption
- Build a model evaluation set
Review prepaid pricing · Read API documentation · Get Started